This invention relates to quality analysis for tube bending processes.
Tubing is used extensively in a variety of situations. One example is exhaust systems for the automotive industry. Tubing manufacture involves numerous processes including cutting, bending, welding and forming. Exhaust tubing is bent to desired shapes for particular vehicle applications using automated equipment that must be capable of high accuracy and repeatability when bending tubing over a wide range of angles. The tubing must be bent without splitting, cracking or collapsing for a wide range of materials (e.g., steel) and thicknesses.
One typical automatic tube bending machine inserts a set of mandrel balls into a tube to support a portion of the tube while that portion is being bent to a desired shape. The mandrel balls are then pulled back automatically by a motor or hydraulic drive mechanism to the next bend position. Occasionally, the mandrel balls will break during the tube bending process as a result of excessive tension or pull force, which may leave mandrel fragments inside the tubing. As a result, mandrel scrap inside the tubing may be assembled into an exhaust system that is later installed onto a vehicle. On occasion this problem has not been discovered until the vehicle has been fully assembled. Troubleshooting problems with the vehicle relating to the presence of mandrel fragments in the exhaust system is difficult and costly. To avoid such problems, OEMs are requiring exhaust system suppliers to incorporate a detection system that ensures that exhaust systems are delivered without mandrel fragments.
One approach to mandrel fragment detection has been to use a camera system to visually detect the presence of mandrel fragments. The use of camera systems has not been that successful since such systems require significant modification to the machine and have resulted in reduction of machine productivity and high maintenance due to the dirty and harsh environments within which the camera systems must perform. Accordingly, a more robust and reliable detection system is needed to determine undesired conditions.